The VMware Renewal Math, Worked Out Loud

The renewal quote lands, it’s several times what it used to be, and the first question is always the same: what does AWS cost?

It’s a reasonable question. It’s also the wrong one to start with, and answering it too fast is how companies end up paying to migrate twice.

We recently priced a real environment for a client weighing exactly this decision: 17 production servers, about 21 TB of file data, a SQL box, a couple of domain controllers, the usual. Here’s what we found, and what it changed about the recommendation.

Where the 20-VM rule comes from

You’ll hear that funded AWS migrations start to make sense somewhere around 20 VMs. That’s roughly true, and worth knowing — but it’s a funding line, not a law of physics.

AWS’s Migration Acceleration Program puts real money behind qualified migrations: assessment, mobilization, and migration support that materially reduce what the move costs. Programs like that have floors. Under the floor, the funding that makes the economics work isn’t there, and a hosted private cloud usually wins. Over it, the math shifts.

So the number matters. It just isn’t the answer — it’s the gate you have to clear before the actual analysis begins.

Compute is cheap. Storage is the bill.

Here’s the part that surprises people. For those 17 servers, priced on a one-year savings plan:

Seventeen Windows and Linux servers running around the clock cost about two grand. The 21 TB of file data sitting behind them — plus keeping a month of daily snapshots — cost nearly twice that. And storage doesn’t discount. Savings plans and reserved capacity cut your compute bill; they do nothing for the terabytes.

That’s not an argument against the cloud. It’s an argument against pricing a lift-and-shift and treating the result as your answer. A straight like-for-like move is the ceiling, not the plan. In that environment, deduplication on a managed file service and moving genuinely cold records to cheap object storage looked like $1,000–1,500 a month — twenty percent of the bill — available before anyone touched a server. A real snapshot retention policy takes more off.

Most environments have the same shape: a third of the data is live, and the rest is records nobody has opened in four years but nobody will authorize deleting. Price that honestly and the comparison changes.

The three questions that actually decide it

1. How much of your data is cold?

If the answer is “most of it,” you have a tiering conversation, not a platform conversation — and you should have it before you compare anything, because it moves the number on both sides.

2. Who operates it at two in the morning?

This is the one that decides most of these, and it rarely shows up on a spreadsheet. The best summary of it we’ve heard came from an IT manager on a call this year, describing a platform his team couldn’t run without help: a Porsche with a tricky clutch, in the shop all the time. He wasn’t arguing about price. He was arguing that a better product he can’t operate is a worse outcome. He was right. If the plan depends on hiring a cloud engineer you aren’t going to hire, it isn’t a plan.

3. What does the company look like in two years?

Acquisitions, an audit, outside investors, a diligence process — these change the answer. Public cloud has a real advantage under scrutiny: auditors check the box and move on. A private environment invites questions, and answering them takes your time. If nothing like that is on the horizon, this doesn’t matter. If it is, it may matter more than the monthly.

You don’t have to choose once

The false premise underneath most of these conversations is that this is a single, permanent decision made under renewal pressure.

It isn’t. A legitimate answer — the one we recommended in the case above — is to move off VMware now onto a hosted private cloud where somebody else carries the operational load, and run the funded AWS assessment in parallel to get a real number instead of an estimate. Then move again in a year or two, deliberately, when the business is ready and the economics are proven rather than projected.

That costs you a second migration. It’s cheaper than it sounds if you design for it: portable disk formats, identity and backup built with the second hop in mind, no dependencies you’ll have to unwind. And it’s considerably cheaper than committing to a platform nobody on your team can run.

Two things not to do

Don’t cancel anything until the destination is proven. The renewal is unpleasant, but an unplanned exit is worse than an expensive quarter.

Don’t manufacture servers to clear a funding threshold. It comes up more than you’d think. It’s the wrong footing to start a multi-year relationship with a cloud provider on, and if your environment genuinely sits under the line, that’s information — not an obstacle.

The honest version

We do this assessment at no cost, and we’ll tell you when the answer is “stay where you are.” We’ve stayed deliberately manufacturer-neutral since 2005 for exactly that reason — the recommendation has to fit you, not our inventory.

Call us at 801-218-3354, or start with your VM count. It’s the first question we ask.